3D Target Detection and Spectral Classification for Single-photon LiDAR Data

TL;DR

This paper introduces a multiscale hierarchical Bayesian approach for 3D surface detection, reconstruction, and spectral classification in single-photon LiDAR data, effectively reducing data volume and enhancing target detection in noisy conditions.

Contribution

It presents a novel multiscale method combined with a hierarchical Bayesian model for efficient 3D reconstruction and spectral classification of LiDAR data, improving robustness against noise.

Findings

Significant data volume reduction achieved

Enhanced target detection in low SNR conditions

Improved 3D reconstruction accuracy

Abstract

3D single-photon LiDAR imaging has an important role in many applications. However, full deployment of this modality will require the analysis of low signal to noise ratio target returns and a very high volume of data. This is particularly evident when imaging through obscurants or in high ambient background light conditions. This paper proposes a multiscale approach for 3D surface detection from the photon timing histogram to permit a significant reduction in data volume. The resulting surfaces are background-free and can be used to infer depth and reflectivity information about the target. We demonstrate this by proposing a hierarchical Bayesian model for 3D reconstruction and spectral classification of multispectral single-photon LiDAR data. The reconstruction method promotes spatial correlation between point-cloud estimates and uses a coordinate gradient descent algorithm for parameter estimation. Results on simulated and real data show the benefits of the proposed target detection and reconstruction approaches when compared to state-of-the-art processing algorithms